A dynamic aggregate model for simulating network-wide travel time reliability

The advancements in Macroscopic Fundamental Diagrams (MFDs) provide opportunities for investigating network-wide travel time reliability. A dynamic aggregated model, combing a Two-Fluid Model (TFM) type MFD, mean-standard deviation relationship of travel time rate and standardized travel time distribution, is proposed. The proposed model only uses taxi GPS data to compute network-wide variables, such as the TFM type MFD, standardized travel time distribution, mean and standard deviation of travel time rate, trip length distribution. Then, the supply side and demand side at the network scale, particularly the network density and departure time choice, can be simulated. Lastly, a Changsha case with taxi GPS data is examined.